It is well established that human skin constantly emits odors which vary with health or environmental conditions. In fact, canines have been trained to detect the changes in skin volatile organic compounds (VOCs) of epileptic and cancer patients. While a powerful tool for health monitoring, skin gas detection remains a largely untouched area due to the hundreds of different gases present at once. With recent advances in machine learning, ‘training’ an array of gas sensors to output a wholistic ‘gas fingerprint’ becomes a possibility. Challenges remain in sensor functionalization to create selective VOC gas detectors, particularly in the presence of humidity. Further, we aim to transition the silicon-based devices to a flexible and wearable sensor array for continuous monitoring.
Project is currently funded by: Federal